Mode Mo Description of the Data

2. Prerequisites Test Analysis

a. Normality

Normality test is used to test the correctness of the data whether it is normally distributed or not. For the purposes of normality testing, this research used Kolmogrov-Smirnov formula. The formula is: Dmax = | Fa x – Fe x | Notes : Dmax = Maximum value of the difference of two cumulative frequency distribution Fa x = Cumulative relative frequency Fe x = Cumulative frequency teorities Djarwanto, 2003: 50 Normality test is used to check whether the data of the investigated population is normally distributed or not. Data considered as normal if the significance value shows the number count of more than 5 or 0.05, or the results of the empulation is less than the score of Kormogrov- Smirnov table.

b. Linearity

Linearity test is intended to determine whether the independent variables and the dependent variable have a linear relationship or not. The formula used in the linearity test in this research as follows: F Notes: F : Score of F number for the regression line RK reg : The quadratic mean of the regression line RK res : The quadratic mean of residual line Sutrisno Hadi, 2004: 13 If obtained the F emp score that is smaller than F table at significance level of 5, the correlation between independent variables and dependent variable is linear. Otherwise, if F emp score is greater than F table , the data is non-linear with a significance level of 5.

c. Multicollinearity

According to Gunawan 2005: 36, multicollinearity testing was intended to prove or test the linear relationship between the independent variable with the other independent variables. To determine whether there is multicollinearity or not in the variables, can use the following formula: . Notes: VIF = standard deviation of inflation factor quadratic Tolerance = the magnitude of the error rate that is given statistically Danang, 2007:89 The limit of tolerance score is 0.1 and the VIF limit is 10. It shows that if the tolerance score is below 0.1 or VIF is above 10, there is a multicollinearity disruption. tolerance VIF 1